Date: Sat, 18 Jan 1997 02:11:36 +0000 From: Brian Somers <brian@awfulhak.demon.co.uk> To: Akihiro Tominaga <tomy@gunpowder.Stanford.EDU> Cc: hackers@freebsd.org Subject: Re: (wide) DHCP negotiation using the REQUEST_IPADDR option Message-ID: <199701180211.CAA04695@awfulhak.demon.co.uk> In-Reply-To: Your message of "Fri, 17 Jan 1997 17:25:51 PST." <199701180125.RAA11735@dynamite.Stanford.EDU>
next in thread | previous in thread | raw e-mail | index | archive | help
> From: Brian Somers <brian@awfulhak.demon.co.uk>
> Subject: Re: (wide) DHCP negotiation using the REQUEST_IPADDR option
> Date: Sat, 18 Jan 1997 00:42:29 +0000
>
> > Right, so we're agreed ? Currently, the code returns -1 (remains silent) if
> > it has a cid & a *different* IP that the one suggested by the client. My
> > patch says it should NAK it.
>
> No, I'm not. I think, you didn't understand what I meant.
>
> Client requests address 'A'.
> There is a server 'X' which has the binding that points address 'B'.
> 'B' is already expired.
> There is another server 'Y' which has the binding that points address 'A'.
> 'A' is valid.
>
> This is not illegal situation, and if 'X' sends NAK and 'Y' sends ACK,
> there is no gurantee which packet arrives first. So the behavior of
> the client changes depending on arrivals of packets.
>
> IMHO, the server should send back NAK only if the DHCPREQUEST is sent
> to the server with Unicast IP address.
>
> I guess, the word 'the server has no **record**' has vagueness.
> I think it should be changed like the following sentence:
>
> the server has non-expired lease
>
> I ask about it at the DHCPv4 ML.
I'm still not sure I understand what you're saying.
In my scenario, the server 'Z' has an explicit (bootp style MAC address)
reference that says IP 10.1.1.1 and the client is asking for IP 10.0.0.1.
IMHO, the server must NAK as it is obviously responsible for telling the
client that it's wrong.
I think you're looking on it as if I've changed the entry from one server
to another - the old server still has a lease, but should be quiet non-the-
less in case it NAKs before the correct server ACKs. The difference is that
the 'Z' server is still responsible for telling the client (in my setup).
If 'Z' doesn't NAK the client, there's no way of ever changing an IP number
until it expires (unless you can tell the client to stop suggesting the old
IP number). This is disasterous if you set up a bad IP for a client, and
boot the client before realizing the mistake.....
--
Brian <brian@awfulhak.demon.co.uk>, <brian@freebsd.org>
<http://www.awfulhak.demon.co.uk/>;
Don't _EVER_ lose your sense of humour....
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199701180211.CAA04695>